Journal of Social and Clinical Psychology, Vol. 30, No. 10, 2011, pp. 1069-1096
Zeigler-Hill et al.
CONTINGENT SELF-ESTEEM
Contingent Self-Esteem and Anticipated
Reactions to Interpersonal Rejection
and Achievement Failure
Virgil Zeigler-Hill
Oakland University
Avi Besser
Sapir Academic College, Israel
Kristi King
University of Southern Mississippi
The anticipated reactivity of individuals with contingent and noncontingent
forms of high self-esteem to imagined self-esteem threats were compared across
two studies using undergraduate participants. The self-esteem threat manipulation in Study 1 (N = 302) involved asking participants to predict their reactions
to discovering that their romantic partner was having a sexual affair, whereas
the manipulation in Study 2 (N = 392) asked participants to consider how they
would respond if they failed to get a promotion that they really wanted at work.
Participants were asked to anticipate their reactions to these scenarios in terms of
state self-esteem, positive affect, negative affect, and anger. Our results revealed
a tendency for individuals with contingent high self-esteem to predict they would
have stronger reactions to these scenarios than individuals with noncontingent
high self-esteem. The pattern of these findings suggests that the protective properties of high self-esteem may be largely limited to individuals who are relatively
secure about their feelings of self-worth.
Address correspondence to Virgil Zeigler-Hill, Department of Psychology, Oakland
University, 111 Pryale Hall, Rochester, MI 48309; E-mail: [email protected]
© 2011 Guilford Publications, Inc.
1069
1070ZEIGLER-HILL ET AL.
Self-esteem refers to the evaluative aspect of self-knowledge that
reflects the extent to which people like themselves (Brown & Marshall, 2006). High self-esteem has been found to be associated with
a range of positive outcomes including subjective well-being (Diener & Diener, 1995) and persistence on difficult tasks (Di Paula &
Campbell, 2002). Not surprisingly, individuals have consistently
been found to show a desire for high self-esteem and are often willing to engage in a variety of strategies to maintain or enhance their
feelings of self-worth (Crocker & Park, 2004). One of the advantages
that is associated with high self-esteem is that it serves as a resource
that protects individuals from potential self-esteem threats such
as rejection or failure (e.g., Beck, 1967; Metalsky, Joiner, Hardin, &
Abramson, 1993; Orth, Robins, & Meier, 2009; Pyszcynski, Greenberg, Solomon, Arndt, & Schimel, 2004). That is, those with high
self-esteem are thought to be less affected by negative experiences
and to recover from these sorts of experiences more quickly than
individuals with low self-esteem. This basic idea has been referred
to using a variety of labels such as the buffer model of high selfesteem and the vulnerability model of low self-esteem (see ZeiglerHill, 2011, for a review). The underlying rationale of models that
emphasize the protective properties of high self-esteem is that negative experiences may be less detrimental for individuals with high
self-esteem because of their enhanced coping resources (Arndt &
Goldenberg, 2002) and the certainty they have regarding their positive characteristics (Campbell, Trapnell, Heine, Katz, Lavallee, &
Lehman, 1996).
The goal of the present study was to gain a more nuanced understanding of the connection between self-esteem and anticipated
reactions to negative events that have the capacity to threaten how
individuals feel about themselves (i.e., interpersonal rejection and
achievement failure). Previous research concerning the protective
properties of high self-esteem has largely focused on self-esteem
level without taking into account other properties of self-esteem.
This focus on the level of self-esteem may explain the conflicted
findings that have emerged concerning the protective properties of
high self-esteem (see Orth et al., 2009, for a review) because high
self-esteem is a heterogeneous construct (e.g., Roberts, 2006; Roberts & Monroe, 1994). That is, it appears that there are subgroups
of individuals with high self-esteem such that some feel genuinely
good about themselves whereas others are defensive and conceited.
These two forms of high self-esteem are often referred to as secure
CONTINGENT SELF-ESTEEM
1071
high self-esteem and fragile high self-esteem, respectively (Kernis,
2003). Individuals with secure high self-esteem have a solid and realistic basis for their feelings of self-worth that does not require constant validation. In contrast, fragile high self-esteem refers to feelings of self-worth that are vulnerable to challenge, require constant
validation, and rely upon some degree of self-deception. As a result,
individuals with fragile high self-esteem are preoccupied with protecting and enhancing their vulnerable feelings of self-worth.
It is possible that self-esteem fragility may moderate the anticipated reactions of high self-esteem individuals to events that may
threaten their feelings of self-worth. More specifically, we expected
that those with fragile high self-esteem would anticipate stronger
reactions to these events than those with secure high self-esteem.
This expected pattern is due to the uncertainty at the core of fragile
self-esteem which may undermine the protective properties of high
self-esteem and leave these individuals vulnerable to the deleterious consequences of self-esteem threats. One marker that is commonly used to distinguish between secure and fragile forms of high
self-esteem is contingent self-esteem. This fragility marker refers to
what an individual believes he or she must do or be in order to have
value and worth as a person (Crocker & Wolfe, 2001; Deci & Ryan,
1995).1 Contingent high self-esteem is considered to be fragile because it can only be maintained as long as the individual is able to
successfully meet the standards upon which his or her self-esteem is
based (Deci & Ryan, 1995; Kernis, 2003). It is important to note that
many of the studies that have focused on contingent self-esteem
have not accounted for its interaction with self-esteem level which
is necessary to capture fragile high self-esteem. However, recent
studies have taken this interaction into account. For example, individuals with contingent high self-esteem (i.e., those who had high
scores on measures of self-esteem level and contingent self-esteem)
have been found to be more defensive than those with noncontingent high self-esteem (e.g., Kernis, Lakey, & Heppner, 2008) and
men with contingent high self-esteem have been found to be more
hostile than other individuals with high self-esteem (Zeigler-Hill,
Clark, & Beckman, in press). It is important to note that researchers
1. We will often refer to contingent and noncontingent forms of self-esteem for
conceptual ease and to avoid confusion with self-esteem level. However, it is important
to note that contingent self-esteem is a continuous variable with contingent and
noncontingent representing the extreme ends of the continuum.
1072ZEIGLER-HILL ET AL.
have also started to examine the interaction of self-esteem level and
domain-specific contingencies which examine the extent to which
individuals base their self-esteem on dimensions such as physical
appearance or academic performance (Park & Crocker, 2005, 2008;
Park, Crocker, & Kiefer, 2007; Park & Maner, 2009).
Overview and Predictions
In order to develop a better understanding of the protective function served by high self-esteem, we examined whether contingent
self-esteem would moderate the association between self-esteem
level and anticipated reactions to negative events that may threaten
one’s feelings of self-worth. Our prediction was that individuals
who reported possessing contingent high self-esteem would anticipate being more reactive to these sorts of potential self-esteem
threats than those with noncontingent high self-esteem. The underlying rationale for our prediction was that individuals with noncontingent high self-esteem may believe they possess the psychological
resources to acknowledge self-esteem threats without being overly responsive to these experiences. The secure and well-anchored
feelings of self-worth that characterize these individuals may lead
them to believe they are capable of confronting their weaknesses
without being threatened by their own lack of perfection. In contrast, the fragile feelings of self-worth that characterize those with
contingent high self-esteem may not be robust enough for these individuals to believe they could incorporate potentially threatening
experiences without being greatly affected by these events. That is,
we expected those with contingent high self-esteem to anticipate
heightened reactivity to imagined self-esteem threats because these
events may have the capacity to undermine their tenuous feelings
of self-worth.
For those individuals with low levels of self-esteem, we predicted that they would anticipate being highly reactive to self-esteem
threats because these individuals lack the coping resources that
characterize those with high self-esteem. However, our predictions
for the combination of low self-esteem and contingent self-esteem
were somewhat uncertain because very little attention has been devoted to the consequences of contingent self-esteem for those with
low levels of self-esteem. We believed the primary difference be-
CONTINGENT SELF-ESTEEM
1073
tween those with contingent and noncontingent forms of low selfesteem would be that those with contingent low self-esteem would
have the opportunity to feel better about themselves on those occasions when they meet the standards that they set for themselves.
However, the present studies focused on negative events that have
the potential to threaten self-esteem which may not offer individuals the opportunity to meet their standards, so the potential benefits
of contingent low self-esteem may not be evident in our studies. As
a result, we expected that both those with contingent and noncontingent forms of low self-esteem would anticipate relatively high
levels of reactivity to the self-esteem threats.
STUDY 1: ANTICIPATED REACTIONS TO SOCIAL REJECTION
The purpose of Study 1 was to examine whether contingent selfesteem moderated the anticipated reactions of individuals with
high self-esteem to an imagined scenario concerning the infidelity
of their romantic partner. We expected individuals with contingent
high self-esteem to anticipate greater reactivity to this manipulation than those with noncontingent high self-esteem because of the
tenuous nature of their feelings of self-worth. We selected romantic
infidelity because self-esteem has been found to be intimately connected with interpersonal relationships (Leary & Downs, 1995) and
past research has shown that individuals are highly reactive to manipulations concerning infidelity (e.g., Besser & Priel, 2009; Besser
& Zeigler-Hill, 2010).
Method
Participants and Procedure
Participants were 302 undergraduates (96 men and 206 women) at a
university in the southern region of the United States. Participants
were enrolled in psychology courses and participated in return for
partial fulfillment of a research participation requirement. Participants were asked to provide written informed consent after the procedures had been fully explained. Although participants were reminded that they could discontinue their participation in the study
at any time, none elected to do so. The mean age of the participants
1074ZEIGLER-HILL ET AL.
was 20.37 years (SD = 3.13) and their racial/ethnic composition was
63% White, 29% Black, and 8% Other.
Participation involved two sessions separated by approximately
one week. During the first session, participants completed measures
of self-esteem level, contingent self-esteem, and other measures that
are not relevant to the present study. Potential order effects were
controlled by presenting the questionnaires in a randomized order.
During the second session, participants were asked to
“Please think of a serious committed romantic relationship that you
currently have, have had in the past, or would like to have in the future. Now that you are thinking about this committed romantic relationship, imagine the following scenario happening . . . ”
Participants were then randomly assigned to read a hypothetical
scenario taken from Besser and Priel (2010) that was intended to
invoke either high or low levels of threat in their romantic relationships. Both scenarios began by describing the following situation:
“You get out of work early and decide to surprise your partner and
buy her/him a present. As you walk up to the apartment, you hear
some laughing coming from inside. As you get closer, you see that the
door is cracked open.”
The scenarios diverge at this point with the high threat scenario
concluding with the following description:
“You open the door to find your partner and another person having
sexual relations in the living room. You hear your partner whispering
to this person, ‘I think I might be in love with you.’”
In contrast, the low threat scenario concludes with this description:
“You open the door to find your partner setting the table while the TV
in the living room shows a couple laughing while they have sexual
relations.’”
After reading the appropriate hypothetical scenario, participants
were asked to report how they believed they would feel if they actually experienced this scenario in terms of their state self-esteem,
positive affect, negative affect, and anger. These outcomes were
selected in order to capture a range of potential responses to the
CONTINGENT SELF-ESTEEM
1075
scenarios with some involving feelings about the self (i.e., state selfesteem) and others dealing with either specific emotional reactions
(i.e., anger) or broad emotional experiences (i.e., negative affect and
positive affect).
Session 1 Measures
Self-Esteem Level. The Rosenberg Self-Esteem Scale (Rosenberg,
1965) is a 10-item measure of global self-esteem (e.g., On the whole,
I am satisfied with myself). Participants were instructed to complete
the instrument according to how they typically or generally feel
about themselves. Responses were made on scales ranging from 1
(strongly disagree) to 5 (strongly agree). This instrument is regarded
as a well-validated and reliable measure of global self-regard (e.g.,
Blaskovich & Tomaka, 1991). The internal consistency of this measure for the present study was α = .88.
Contingent Self-Esteem. The Contingent Self-Esteem Scale (Paradise & Kernis, 1999) is a 15-item measure of general self-esteem
contingency (e.g., When my actions do not live up to my expectations, it makes me feel dissatisfied with myself). Participants were
asked to respond to these items on scales ranging from 1 (not at all
like me) to 5 (very much like me). The Contingent Self-Esteem Scale
has been found to be a reliable and valid measure of contingent selfesteem (Kernis & Goldman, 2006). The internal consistency of this
measure for the present study was α = .79.
Session 2 Measures
State Self-Esteem. The State Self-Esteem Scale (Heatherton & Polivy, 1991) is a 20-item measure that assesses self-esteem for the
following three domains: performance (e.g., I feel confident about
my abilities), social (e.g., I feel concerned about the impression I
am making [reverse-scored]), and appearance (e.g., I feel satisfied
with the way my body looks right now). Participants were asked to
complete this measure based on how they believed they would feel
about themselves if they experienced the event described in the social rejection scenario. Responses were made on scales ranging from
1 (not at all) to 9 (extremely). Previous research has shown the State
Self-Esteem Scale to be a valid and reliable measure of self-esteem
(e.g., Heatherton & Polivy, 1991; Vohs & Heatherton, 2004). The
1076ZEIGLER-HILL ET AL.
internal consistency of the composite measure of state self-esteem
was .94.2
Affect. Affect was measured using the Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988) which is
a reliable and well-validated self-report measure of this construct.
The PANAS consists of scales that measure positive (e.g., interested, enthusiastic, proud) and negative affect (e.g., distressed, scared,
hostile). Participants were instructed to complete the items according to how they believed they would feel if they experienced the
event described in the scenario. Responses for each of the 20 items
were made on scales ranging from 1 (very slightly or not at all) to
9 (extremely). For the present sample, the internal consistencies of
these scales were .87 and .92 for positive affect and negative affect,
respectively.
Anger. Anger was measured using the state anger scale of the
State-Trait Anger Expression Inventory-2 (STAXI-2; Spielberger,
1999). The STAXI-2 consists of 15 items that capture temporary feelings of anger (e.g., I feel annoyed). Participants were instructed to
complete the items according to how they believed they would feel
if they experienced the event described in the scenario. Responses
for each of the items were made on scales ranging from 1 (not at all)
to 9 (very much so). For the present sample, the internal consistency
of the STAXI-2 was α = .98.
Results
The means, standard deviations, and intercorrelations for the measures in Study 1 are presented in Table 1. Analyses were conducted
to determine whether self-esteem level and contingent self-esteem
moderated the anticipated reactions of participants to the infidelity
scenario (i.e., social rejection threat). The associations that the potential moderator (i.e., contingent self-esteem) had with the predictor variables (i.e., self-esteem level and condition) were relatively
small. As a result, it is reasonable to examine whether contingent
2. Preliminary analyses also examined each of the state self-esteem subscales (i.e.,
performance self-esteem, social self-esteem, and appearance self-esteem). The results of
those analyses were similar to what emerged for the total composite score. As a result,
we trimmed the analyses concerning the subscales from the manuscript in the interest
of parsimony.
CONTINGENT SELF-ESTEEM
1077
self-esteem moderates the associations that the predictor variables
have with the anticipated reactions of participants to the scenarios
(Jaccard, Turrisi, & Wan, 1990). This was accomplished by conducting a series of hierarchical multiple regression analyses in which
each indicator of reactivity (i.e., state self-esteem, positive affect,
negative affect, and anger) was regressed onto threat condition (0
= low threat, 1 = high threat), self-esteem level, contingent selfesteem, and sex (0 = female, 1 = male). We included sex in these
analyses because it has been found to moderate the association that
fragile self-esteem has with outcomes such as interpersonal style
(Zeigler-Hill et al., in press) and aggression (Webster, Kirkpatrick,
Nezlek, Smith, & Paddock, 2007). Preliminary analyses found that
sex did not moderate any of the results presented in later sections
so the interaction terms involving sex were trimmed from the final models in the interest of parsimony.3 The continuous predictor
variables were standardized for the purpose of testing interactions
(Aiken & West, 1991). For these analyses, the main effect terms for
threat condition, self-esteem level, contingent self-esteem, and sex
were entered on Step 1. The two-way interactions of threat condition, self-esteem level, and contingent self-esteem were entered on
Step 2 and their three-way interaction was entered on Step 3. The
results of these analyses are presented in Table 2. These regression
analyses were followed by the simple slopes tests recommended
by Aiken and West (1991) to describe the interaction of continuous
variables.
State Self-Esteem. The main effect emerged for condition (β = -.37,
t = -7.04, p < .001, d = -.83) such that those who experienced social
rejection threat anticipated lower state self-esteem than those in the
low threat condition. Main effects also emerged for self-esteem level
(β = .11, t = 2.13, p < .05, d = .25) and contingent self-esteem (β =
-.15, t = -2.77, p < .01, d = -.32) such that those with higher levels of
self-esteem and lower levels of contingent self-esteem thought they
would experience higher state self-esteem. The main effects of condition and contingent self-esteem were qualified by their interaction (β = -.20, t = -2.29, p < .05, d = -.27). The predicted values for this
3. We also included a term in the preliminary analyses concerning relationship status
(0 = currently single, 1 = in a committed relationship) in order to distinguish between
those participants who were involved in a romantic relationship and those who were
not. However, no main effects or interactions involving relationship status emerged
from our analyses so the term was trimmed from the final analyses.
1078ZEIGLER-HILL ET AL.
interaction are presented in Panel A of Figure 1. Simple slopes tests
found that the slope of the line representing the association between
social rejection threat and state self-esteem was negative for those
with contingent self-esteem (β = -.50, t = -6.40, p < .001, d = -.75) as
well as those with noncontingent self-esteem (β = -.24, t = -3.24, p <
.001, d = -.38). However, it is important to note that the association
between social rejection threat and anticipated state self-esteem was
significantly stronger for those with contingent self-esteem such
that individuals with contingent self-esteem anticipated significantly lower levels of state self-esteem than those with noncontingent
self-esteem in the high threat condition (β = -.26, t = -3.41, p < .001,
d = -.40). Taken together, these results show that the lowest levels of
state self-esteem were anticipated by individuals with contingent
self-esteem who imagined experiencing a social rejection threat. In
contrast, the anticipated state self-esteem levels of those with high
levels of contingent self-esteem did not differ from those with low
levels of contingent self-esteem in the absence of threat (β = .02, t <
1, ns).
Positive Affect. The results of the analysis concerning positive affect revealed a main effect for threat condition (β = -.33, t = -6.17, p
< .001, d = -.72) such that those participants in the high threat condition anticipated less positive affect than those in the low threat
condition. In addition, main effects emerged for self-esteem level
(β = -.14, t = -2.57, p < .05, d = -.30) and contingent self-esteem (β =
-.16, t = -2.77, p < .01, d = -.32). No other main effects or interactions
emerged from this analysis.
Negative Affect. Main effects emerged for threat condition (β = .51,
t = 10.51, p < .001, d = 1.23) and contingent self-esteem (β = .21, t
= 4.04, p < .001, d = .47) from the analysis concerning anticipated
negative affect but these main effects were qualified by the expected
three-way interaction of threat condition, self-esteem level, and contingent self-esteem (β = .24, t = 2.69, p < .01, d = .31). The predicted
values for this interaction are presented in Panel B of Figure 1. As
suggested by Cohen, Cohen, West, and Aiken (2003), this interaction
was probed by first examining whether the two-way interaction of
condition and contingent self-esteem was significant for those with
low and high levels of self-esteem. These analyses found that this
two-way interaction emerged for those with high self-esteem (β =
.79, t = 3.79, p < .001, d = .45) but not for those with low self-esteem
(β = .18, t = 1.55, ns). Simple slopes tests were then conducted which
4.10
0.68
3.94
0.76
.33***
-.62***
—
.29***
-.69***
-.52***
-.23***
—
.07
.16*
.01
-.18*
5.69
1.35
5.19
1.50
.18*
-.31***
—
.28***
-.86***
-.75***
6.21
1.42
4.84
1.77
-.23**
—
-.07
-.16
-.10
-.16
3.35
0.54
3.52
0.61
3.39
0.48
3.34
0.44
3
2
5.62
1.74
4.26
1.56
-.16*
-.17*
.19**
—
.09
-.02
-.22**
-.15
.31***
—
-.09
-.11
4.70
1.60
3.52
1.51
4
3.81
1.84
4.44
1.65
-.34***
.48***
-.78***
.14*
—
.79***
-.10
.48***
-.65***
-.19*
—
.82***
3.30
1.73
5.55
1.82
5
2.69
1.88
4.00
2.02
-.35***
.41***
-.59***
.17*
.79***
—
-.03
.05
-.24**
-.35***
.53***
—
2.86
1.81
7.14
2.14
6
Note. Correlations for those in the low threat conditions are presented below the diagonals while correlations for the high threat conditions are presented above the diagonals. *p < .05; **p < .01; ***p < .001.
MLow Threat
SDLow Threat
MHigh Threat
SDHigh Threat
1
Study 1: Anticipated Reactions to Social Rejection
1. Self-Esteem Level
—
2. Contingent Self-Esteem
-.17*
3. State Self-Esteem
.13
4. Positive Affect
.03
5. Negative Affect
.02
6. Anger
-.03
MLow Threat
4.04
0.67
SDLow Threat
4.05
MHigh Threat
0.78
SDHigh Threat
Study 2: Anticipated Reactions to Achievement Failure
1. Self-Esteem Level
—
2. Contingent Self-Esteem
-.38***
3. State Self-Esteem
.23**
4. Positive Affect
.00
5. Negative Affect
-.11
6. Anger
-.06
Table 1. Intercorrelations and Descriptive Statistics for Measures of Self-Esteem Level, Contingent Self-Esteem,
and Anticipated Reactions to Self-Esteem Threat
CONTINGENT SELF-ESTEEM
1079
Step 1
.02*
Condition x SEL x CSE
.01
-.07
.04
SEL x CSE
.25***
-.20*
Condition x CSE
Step 3
-.03
.24***
Condition x SEL
Step 2
.10
-.15**
Contingent Self-Esteem (CSE)
Sex
.11*
.20***
.20***
.18***
.00
.02*
.18***
β
.06
.09
-.01
-.04
.07
-.16**
-.14*
-.33***
.22***
.22***
-.37***
∆R2
Self-Esteem Level (SEL)
Condition
Positive Affect
R2
∆R2
β
State Self-Esteem
R2
Study 1: Anticipated Reactions to Social Rejection
.41***
.39***
.02*
.06***
.33***
∆R2
β
.24**
.00
.38***
.00
.03
.21***
.00
.51***
Negative Affect
.33***
R2
.58***
.57***
.57***
R2
Table 2. Regressions of the State Self-Esteem, Affect, and Anger on Self-Esteem Threat Condition,
Self-Esteem Level, Contingent Self-Esteem, and Sex
.01
.00
.57***
∆R2
Anger
.14
-.03
.08
.02
.11**
.00
-.03
.74***
β
1080ZEIGLER-HILL ET AL.
Step 3
.33***
.04**
-.14*
2
2
Note. ∆R = increase in R . *p < .05; **p < .01; ***p < .001.
Condition x SEL x CSE
-.06
-.45***
SEL x CSE
Condition x CSE
.10***
-.03
Step 2
Condition x SEL
.29***
-.28***
Contingent Self-Esteem (CSE)
.07
.21***
.19***
-.17***
.19***
Self-Esteem Level (SEL)
Sex
R
.21***
.20***
.16***
.01
.04***
.16***
β
.03
.08
-.20***
-.03
-.09
-.05
-.10*
-.39***
∆R
β
2
∆R
2
Positive Affect
2
State Self-Esteem
Condition
Step 1
R
2
Study 2: Anticipated Reactions to Achievement Failure
.22***
.20***
.02*
.09***
.11***
∆R
2
β
.12*
.08
.23***
-.03
-.05
.17**
-.18***
.17***
Negative Affect
.11***
R
2
.27***
.27***
.15***
R
2
.00
.12***
.15***
∆R2
Anger
.06
.07
.30***
-.04
-.02
.06
-.20***
.30***
β
CONTINGENT SELF-ESTEEM
1081
1082ZEIGLER-HILL ET AL.
Figure 1. Study 1: Predicted values for state self-esteem (Panel A) and
negative affect (Panel B) illustrating the interaction of threat condition,
self-esteem level, and contingent self-esteem at values that are one
standard deviation above and below their respective means.
found that the slope of the line representing the association between
threat condition and anticipated negative affect was significant for
individuals with contingent high self-esteem (β = .87, t = 6.19, p <
.001, d = .73) but not for other individuals (β < .26, ts < 1.05, ns).
These results show that social rejection threat had the greatest impact on the anticipated negative affect of those with contingent high
self-esteem.
Anger. The main effect of threat condition (β = .74, t = 18.71, p <
.001, d = 2.18) emerged from the analysis concerning anger such that
those in the high threat condition anticipated higher levels of anger
than those in the low threat condition. The main effect of gender
CONTINGENT SELF-ESTEEM
1083
also emerged (β = .11, t = 2.64, p < .01, d = .31) such that men reported anticipating higher levels of anger than women. The expected
three-way interaction of condition, self-esteem level, and contingent
self-esteem approached—but did not reach—conventional levels of
significance (β = .14, t = 1.82, p < .07, d = .21). The predicted values
for this analysis revealed a pattern that was similar to the results for
negative affect such that the social rejection threat had the greatest
impact on the anticipated anger of those with contingent high selfesteem.
Discussion
The results of Study 1 provided partial support for our predictions.
That is, the predicted three-way interaction of condition, self-esteem
level, and contingent self-esteem emerged for negative affect such
that those with contingent high self-esteem anticipated being more
responsive to the social rejection threat than those with noncontingent high self-esteem. Although the expected pattern emerged for
negative affect, it failed to emerge for the other outcomes examined
in this study. Despite the fact that the expected three-way interaction failed to emerge for state self-esteem, further exploration of
these results found that the interaction of condition and contingent
self-esteem was only significant for those with high levels of selfesteem. Thus, the expected pattern may have simply been too weak
to rise to conventional levels of significance. This is not terribly
surprising given the difficulty that has been noted with detecting
complex interactions involving continuous variables (McClelland
& Judd, 1993).
It is possible that our difficulty in capturing the expected threeway interaction for some of the outcomes we examined was due to
the social rejection manipulation being so powerful. In other words,
asking participants to imagine themselves walking in on their romantic partners having sex with someone else—and hearing them
whisper that they might be in love with this other person—may
have been so unpleasant and aversive that it may have partially
masked the differences between contingent and noncontingent
forms of high self-esteem. Due to this concern, we used a somewhat
weaker manipulation in Study 2.
1084ZEIGLER-HILL ET AL.
STUDY 2: ANTICIPATED REACTIONS
TO ACHIEVEMENT FAILURE
Our goal for Study 2 was to replicate the findings of Study 1 and to
extend these findings to the domain of achievement. More specifically, we examined whether individuals with contingent high selfesteem anticipated stronger reactions to achievement failure than
those with noncontingent high self-esteem. In addition to shifting
the domain from social rejection to achievement failure, we also
wanted to utilize a manipulation that was less powerful than the
infidelity manipulation employed in Study 1 in order to make it less
likely that the aversive nature of the manipulation would mask any
differences between those with contingent and noncontingent forms
of high self-esteem. As a result, we decided to use the achievement
failure manipulation employed by Besser and Priel (2010) because
they found that it elicited weaker reactions than the infidelity manipulation we used in Study 1. It is important to note that other
studies using similar scenarios concerning interpersonal rejection
and achievement failure have also found that interpersonal rejection tends to elicit stronger reactions than achievement failure (e.g.,
Besser & Zeigler-Hill, 2010).4
Method
Participants and Procedure
Participants were 392 undergraduates (132 men and 260 women).
The mean age of the participants was 20.98 years (SD = 3.84) and
their racial/ethnic composition was 61% White, 29% Black, and 10%
Other. The procedure for Study 2 was very similar to that of Study
1 (e.g., two sessions separated by approximately one week). The fo-
4. As expected, comparisons of the social rejection and achievement failure
conditions across our two studies found that participants in the social rejection
condition reported lower state self-esteem (t = 2.09, p < .05, d = .21), less positive affect (t
= 4.57, p < .001, d = .48), more negative affect (t = 6.11, p < .001, d = .64), and more anger
(t = 14.43, p < .001, d = 1.51) than those in the achievement failure condition.
CONTINGENT SELF-ESTEEM
1085
cus of the second session was on anticipated reactions to a threatening achievement event such that participants were asked to
“Please think of a serious long-term job that you currently have,
have had in the past, or would like to have in the future. Now that
you are thinking about this job, imagine the following scenario
happening . . . .”
Participants were then randomly assigned to read a hypothetical
scenario taken from Besser and Priel (2010) that was intended to
invoke either high or low levels of threat concerning achievement
failure. Both scenarios began by describing the following situation:
“Recently an opportunity for promotion is opened for one exceptional
employee at your place of work. You are competing for this opportunity and you really want this promotion. You have been invited to a
meeting with the executive manager. You approach the executive manager’s office earlier than expected. As you walk up to the office, you
hear some laughing coming from inside. It seems they are celebrating—they probably already know who won the promotion. As you get
closer, you see that the door is cracked open.”
The scenarios diverge at this point with the high threat scenario
concluding with the following description:
“You open the door, to find the executive manager making a toast with
one of your fellow employees to celebrate his promotion. You hear the
executive manager saying to this other employee, ‘Of all the candidates for this promotion, you were clearly the best.’”
In contrast, the low threat scenario concludes with this description:
“You open the door to find the executive manager making a toast with
his secretary who is about to retire. You hear the executive manager saying to her, ‘Thanks for your highly professional work over the
years.’”
After reading the appropriate hypothetical scenario, participants
were asked to report how they believed they would feel if they actually experienced this scenario.
1086ZEIGLER-HILL ET AL.
Session 1 Measures
The same measures from Study 1 were used to assess self-esteem
level (α = .90), and contingent self-esteem (α = .76).
Session 2 Measures
The same measures from Study 1 were used to assess anticipated
reactions to the scenarios in terms of state self-esteem (α = .86), positive affect (α = .94), negative affect (α = .90), and anger (α = .97).
Results
The means, standard deviations, and intercorrelations for the measures in Study 2 are presented in Table 1. Hierarchical multiple
regression analyses similar to those used in Study 1 were used to
examine whether contingent self-esteem moderated the association
between self-esteem level and anticipated reactions of individuals
to the achievement failure manipulation in Study 2. The correlations
that contingent self-esteem had with self-esteem level and condition were relatively modest which allows us to examine contingent
self-esteem as a potential moderator. The results of these analyses
are presented in Table 2.
State Self-Esteem. The main effect emerged for condition (β = -.17,
t = -3.59, p < .001, d = -.37) such that those who experienced the
achievement failure threat anticipated lower state self-esteem than
those in the low threat condition. Main effects also emerged for selfesteem level (β = .21, t = 4.33, p < .001, d = .44), and contingent selfesteem (β = -.28, t = -5.52, p < .001, d = -.56) such that higher levels
of anticipated state self-esteem were reported by individuals with
higher levels of self-esteem and individuals with lower levels of
contingent self-esteem. The main effects of condition, self-esteem
level, and contingent self-esteem were qualified by the expected
three-way interaction (β = -.14, t = -2.14, p < .05, d = -.22). The predicted values for this interaction are presented in Panel A of Figure
2. This three-way interaction was probed by first examining whether the two-way interaction of condition and contingent self-esteem
would emerge for those with low and high levels of self-esteem.
These analyses found that this two-way interaction emerged for
Figure 2. Study 2: Predicted values for state self-esteem (Panel A), positive affect (Panel
B), negative affect (Panel C), and anger (Panel D) illustrating the interaction of threat
condition, self-esteem level, and contingent self-esteem at values that are one standard
deviation above and below their respective means.
CONTINGENT SELF-ESTEEM
1087
1088ZEIGLER-HILL ET AL.
those with high self-esteem (β = -.56, t = -6.85, p < .001, d = -.70) as
well as for those with low self-esteem (β = -.29, t = -2.88, p < .01, d =
-.29). Simple slopes tests were then conducted which found that the
slope of the line representing the association between threat condition and state self-esteem was negative for individuals with contingent high self-esteem (β = -.60, t = -6.19, p < .001, d = -.63) and contingent low self-esteem (β = -.38, t = -4.91, p < .001, d = -.50). In contrast,
the association between threat condition and state self-esteem was
actually positive for those with noncontingent high self-esteem (β =
.20, t = 2.72, p < .01, d = .28). The association between threat and state
self-esteem was not significant for those with noncontingent low
self-esteem (β = .03, t < 1, ns). Taken together, these results show that
individuals with contingent high self-esteem anticipated the greatest drop in their state self-esteem following achievement failure. It
is important to note that individuals with noncontingent high selfesteem actually anticipated a slight increase in their feelings of selfworth following achievement failure.
Positive Affect. Main effects emerged for threat condition (β = -.39,
t = -8.32, p < .001, d = .84) and self-esteem level (β = -.10, t = -2.02,
p < .05, d = .20). The main effect of threat condition was moderated
by its interaction with contingent self-esteem (β = -.20, t = -3.98, p <
.001, d = .40). The predicted values for the interaction of threat condition and contingent self-esteem are presented in Panel B of Figure
2. Simple slopes tests found that the slopes of the lines representing
the associations between achievement failure and anticipated positive affect were negative for those with contingent self-esteem (β =
-.59, t = -8.86, p < .001, d = .90) as well as those with noncontingent
self-esteem (β = -.20, t = -2.95, p < .01, d = .30) but the association
was stronger for those with contingent self-esteem. Taken together,
these results show that achievement failure was associated with anticipated positive affect for all participants but the association was
stronger for those with contingent self-esteem.
Negative Affect. Main effects for threat condition (β = .17, t = 3.46, p
< .001, d = .35), self-esteem level (β = -.18, t = -3.50, p < .001, d = .35),
and contingent self-esteem (β = .17, t = 3.17, p < .01, d = .32) emerged
from the analysis concerning anticipated negative affect but these
main effects were qualified by their three-way interaction (β = .12, t
= 2.41, p < .05, d = .24). The predicted values for this interaction are
presented in Panel C of Figure 2. This three-way interaction was
probed by first examining whether the two-way interaction of con-
CONTINGENT SELF-ESTEEM
1089
dition and contingent self-esteem would emerge for those with low
and high levels of self-esteem. These analyses found that this twoway interaction emerged for those with high self-esteem (β = .50, t
= 3.48, p < .001, d = .35) but not for those with low self-esteem (β =
-.20, t = -1.10, ns). Simple slopes tests were then conducted which
found that the slope of the line representing the association between
achievement failure condition and anticipated negative affect was
significant for individuals with contingent high self-esteem (β = .37,
t = 4.33, p < .001, d = .44) but not for other individuals (βs < .17, ts <
1.01, ns). Taken together, these results show that achievement failure had the greatest impact on the levels of negative affect that were
expected by those with contingent high self-esteem.
Anger. Main effects emerged for threat condition (β = .30, t = 6.32,
p < .001, d = .64) and self-esteem level (β = -.20, t = -3.91, p < .001,
d = .40) from the analysis concerning anticipated anger. The main
effect of threat condition was moderated by its interaction with
contingent self-esteem (β = .30, t = 6.37, p < .001, d = .65). The predicted values for this interaction are presented in Panel D of Figure
2. Simple slopes tests found that the slope of the line representing
the association between achievement failure and anticipated anger
was positive for those with contingent self-esteem (β = .58, t = 9.13,
p < .001, d = .92) but not for those with noncontingent self-esteem
(β = -.01, t < 1, ns). These results show that achievement failure is
associated with the expectation of heightened anger for those with
contingent self-esteem but not for those with noncontingent selfesteem.
Discussion
The results of Study 2 provided partial support for our predictions.
The expected three-way interaction emerged for state self-esteem
and negative affect which showed that individuals with contingent
high self-esteem thought they would be more reactive to achievement failure than those with noncontingent high self-esteem. In addition, individuals with contingent self-esteem anticipated stronger
reactions to this threat manipulation than those with noncontingent
self-esteem for each of the remaining outcomes. Importantly, further
probing of the results found that the interaction between condition
and contingent self-esteem was only significant for those with high
self-esteem for all of the outcomes even if the three-way interaction
1090ZEIGLER-HILL ET AL.
did not reach conventional levels of significance. As with Study 1,
this pattern provides consistent support for our prediction even if
the higher-order interactions were too weak to be detected in some
cases.
General Discussion
The purpose of the present studies was to examine whether contingent self-esteem moderated the association between self-esteem
level and anticipated reactions to events with the potential to threaten how individuals feel about themselves. Across two studies, we
found support for our prediction that individuals with contingent
high self-esteem would anticipate having stronger reactions to social rejection and achievement failure than those with noncontingent high self-esteem. The expected pattern emerged most clearly
for anticipated state self-esteem and negative affect which were
closely associated with each other. That is, individuals who thought
they would experience high levels of negative affect if they experienced the situations described in the scenarios also tended to predict that they would experience lower levels of state self-esteem
in those situations. This suggests that individuals believe that the
state self-esteem they will experience is more closely aligned with
their anticipated negative affect than the other potential reactions
that were examined in the present studies (i.e., positive affect and
anger). The pattern of these results suggests at least some degree
of specificity in the protective properties offered by noncontingent
high self-esteem.
It is important to note that this pattern was clearer for the achievement failure manipulation than it was for the social rejection manipulation. The reason for the difference in how individuals responded
to these manipulations may be that the social rejection manipulation was so powerful that it threatened the self-esteem of those with
both contingent and noncontingent forms of high self-esteem. The
aversive nature of the social rejection manipulation may also explain
why individuals with noncontingent self-esteem anticipated more
positive affect and less anger than those with contingent self-esteem
following the achievement failure manipulation even though no
differences emerged between these individuals following the social
rejection manipulation.
CONTINGENT SELF-ESTEEM
1091
These results may have implications for the choices that are made
by individuals with contingent and noncontingent forms of high
self-esteem. For example, individuals with contingent high selfesteem may be more conservative than those with noncontingent
high self-esteem with regard to the choices they make during the
course of their lives due to their belief that they will experience
strong emotional reactions to aversive events. As a result, individuals with contingent high self-esteem may be less willing to take
risks such as asking for a promotion at work because they may be
concerned about how they would react if their request was denied.
In contrast, individuals with noncontingent high self-esteem may
be more willing to pursue their goals and take risks because of their
belief that they will be able to cope with failure or rejection. These
results complement those of Park (2010) concerning the connection
that self-esteem has with approach and avoidance motivation.
Although the present study focused on the reactions that individuals believed they would experience when confronted with
self-esteem threats, these results suggest that it may be important
to account for contingent self-esteem when considering the reactions of individuals to actual negative events that they experience
during the course of their lives. This argument is similar to the one
presented by Roberts and Monroe (1994) concerning self-esteem
instability (i.e., fluctuations in moment-to-moment feelings of selfworth). The pattern of results observed across the present studies
suggests that the protective properties of high self-esteem may be
limited to those individuals who feel secure about their feelings of
self-worth. These results may shed additional light on past inconsistencies which found that high levels of self-esteem protected individuals from negative experiences in some studies but failed to do
so in other studies. More specifically, individuals with secure high
self-esteem may experience the protective properties of high selfesteem whereas those with fragile high self-esteem may not have
the benefit of that protection.
Contingent self-esteem requires at least some level of awareness
concerning the fragile nature of their feelings of self-worth. This is
necessary because contingent self-esteem is assessed through direct
self-reports which require individuals to have adequate insight concerning their feelings of self-worth to recognize that they are contingent. This differentiates contingent self-esteem from the other
markers of fragile high self-esteem (i.e., low implicit self-esteem
and unstable self-esteem) which do not necessarily involve this
1092ZEIGLER-HILL ET AL.
sort of insight. For example, individuals with unstable high selfesteem are not necessarily aware that their self-esteem is changing
so frequently or able to recognize the sorts of events that lead to
these fluctuations (e.g., Kernis et al., 2008). The insight required for
contingent self-esteem may be important given the manipulations
used in the present studies required participants to predict how
they would respond if they experienced certain events. It is possible
that individuals who are identified as possessing fragile high selfesteem using the other markers may be less aware of their vulnerability to self-esteem threats than those with contingent self-esteem.
It would be informative for future research to replicate these studies
using the other markers of fragile high self-esteem to gain a better understanding of the similarities and differences between these
markers. Also, future studies may want to extend these results to
domain-specific contingencies such as those developed by Crocker
and her colleagues (e.g., Park & Crocker, 2005).
The present studies had a number of strengths which include
their use of large samples, manipulations of self-esteem threat in
two domains, and multiple outcome measures to assess different
aspects of anticipated reactivity. Despite the strengths of these studies, it is also important to acknowledge some of their limitations.
First, we relied on individuals imagining negative experiences as
well as predicting their reactions to these sorts of events. As a result, we know how individuals say they would react to these sorts
of events but this may not necessarily be an accurate reflection of
how they would actually respond. Future researchers may want to
consider using manipulations that avoid this problem by exposing
individuals to actual self-esteem threats in the laboratory. For example, researchers have used laboratory manipulations concerning
mood manipulations (Kernis, Greenier, Herlocker, Whisenhunt, &
Abend, 1997) and social approval/disapproval (Leary et al., 2003)
as well as naturalistic experiences such as earning good/bad grades
(Crocker, Karpinski, Quinn, & Chase, 2003) or receiving acceptance/rejection letters from graduate schools (Crocker, Sommers,
& Luhtanen, 2002). Second, we were unable to determine whether
contingent high self-esteem actually causes heightened reactivity
due to the correlational nature of our data. The process model that
guided our research was that individuals with contingent high selfesteem would lack the sort of protection that noncontingent high
self-esteem provides but this cannot be established using the present data. For example, it is unclear whether contingent high self-
CONTINGENT SELF-ESTEEM
1093
esteem causes individuals to anticipate being highly reactive to selfesteem threats or if the direction of causation was reversed such
that the tendency to anticipate greater reactivity may contribute to
the development of contingent high self-esteem. Further research is
needed to gain a clearer understanding of the causal link between
contingent self-esteem and reactivity. Third, the present studies
relied exclusively on self-report measures which make it possible
that our results may be influenced by socially desirable response
distortions. For example, the same individuals who denied possessing contingent self-esteem may have also claimed that they would
not be upset if they experienced these sorts of events. Fourth, the
generalizability of the present findings may be limited due to our
reliance on undergraduate participants. Undergraduates have most
likely had only limited experience with romantic relationships and
occupational achievement, so their beliefs about how they would
respond to these sorts of self-esteem threats may be somewhat different from older individuals who have more experience in these
areas. Fifth, we relied on a single threat scenario for each study (i.e.,
infidelity and failure to get a promotion). As a result, we only have
a narrow sampling of stimuli which limits our ability to generalize beyond these particular scenarios to the broader issues of social
rejection and achievement failure (e.g., Wells & Windschitl, 1999).
Sixth, we did not collect pre-manipulation measures of affect or anger. This is a limitation because it prevents us from assessing the
extent to which participants would have reported changes in their
affect and anger following the manipulation.
Conclusion
The findings of the present studies suggest that individuals with
contingent high self-esteem anticipate being more reactive to threat
than those with noncontingent high self-esteem. That is, contingent
high self-esteem was associated with predictions concerning stronger emotional reactions following self-esteem threats. These findings suggest the intriguing possibility that the protective properties
of high self-esteem may be limited to those with the secure form of
high self-esteem. These results extend our understanding of the link
between fragile high self-esteem and anticipated reactions to negative events in diverse domains.
1094ZEIGLER-HILL ET AL.
References
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. London: Sage.
Arndt, J., & Goldenberg, J. L. (2002). From threat to sweat: The role of physiological
arousal in the motivation to maintain self-esteem. In A. Tesser, D. A. Stapel,
& J. V. Wood (Eds.), Self and motivation: Emerging psychological perspectives (pp.
43-69). Washington, DC: American Psychological Association.
Beck, A. T. (1967). Depression: Clinical, experimental, and theoretical aspects. New York:
Harper & Row.
Besser, A., & Priel, B. (2009). Emotional responses to a romantic partner’s imaginary
rejection: The roles of attachment anxiety, covert narcissism, and self-evaluation. Journal of Personality, 77, 287-325.
Besser, A., & Priel, B. (2010). Grandiose narcissism versus vulnerable narcissism in
threatening situations: Emotional reactions to achievement failure and interpersonal rejection. Journal of Social and Clinical Psychology, 29, 874-903.
Besser, A., & Zeigler-Hill, V. (2010). The influence of pathological narcissism on
emotional and motivational responses to negative events: The roles of visibility and concern about humiliation. Journal of Research in Personality, 44,
520-534.
Blaskovich, J., & Tomaka, J. (1991). Measures of self-esteem. In J. P. Robinson, P. R.
Shaver, & L. S. Wrightsman (Eds.), Measures of personality and social psychological attitudes (Vol. 1, pp. 115-160). New York: Academic.
Brown, J. D., & Marshall, M. A. (2006). The three faces of self-esteem. In M. H. Kernis
(Ed.), Self-esteem issues and answers: A source book of current perspectives (pp.
4-9). New York: Psychology Press.
Campbell, J. D., Trapnell, P. D., Heine, S. J., Katz, I. M., Lavallee, L. F., & Lehman,
D. R. (1996). Self-concept clarity: Measurement, personality correlates, and
cultural boundaries. Journal of Personality and Social Psychology, 70, 141-156.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the behavioral sciences (3rd ed.). Mahwah, NJ: Erlbaum.
Crocker, J., Karpinski, A., Quinn, D. M., & Chase, S. K. (2003). When grades determine self-worth: Consequences of contingent self-worth for male and female
engineering and psychology majors. Journal of Personality and Social Psychology, 85, 507-516.
Crocker, J., & Park, L. E. (2004). The costly pursuit of self-esteem. Psychological Bulletin, 130, 392-414.
Crocker, J., Sommers, S. R., & Luhtanen, R. K. (2002). Hopes dashed and dreams
fulfilled: Contingencies of self-worth and admissions to graduate school. Personality and Social Psychology Bulletin, 28, 1275-1286.
Crocker, J., & Wolfe, C. T. (2001). Contingencies of self-worth. Psychological Review,
108, 593-623.
Deci, E. L., & Ryan, R. M. (1995). Human agency: The basis for true self-esteem.
In M. H. Kernis (Ed.), Efficacy, agency, and self-esteem (pp. 31-50). New York:
Plenum Press.
Diener, E., & Diener, M. (1995). Cross-cultural correlates of life satisfaction and selfesteem. Journal of Personality and Social Psychology, 68, 653-663.
Di Paula, A., & Campbell, J.D. (2002). Self-esteem and persistence in the face of failure. Journal of Personality and Social Psychology, 83, 711-724.
CONTINGENT SELF-ESTEEM
1095
Heatherton, T. F., & Polivy, J. (1991). Development and validation of a scale for
measuring state self-esteem. Journal of Personality and Social Psychology, 60,
895-910.
Jaccard, J., Turrisi, R., & Wan, C. K. (1990). Interaction effects in multiple regression.
Beverly Hills, CA: Sage.
Kernis, M. H. (2003). Toward a conceptualization of optimal self-esteem. Psychological Inquiry, 14, 1-26.
Kernis, M. H., & Goldman, B. M. (2006). Assessing stability of self-esteem and contingent self-esteem. In M. H. Kernis (Ed.), Self-esteem issues and answers: A
source book of current perspectives (pp. 77-85). New York: Psychology Press.
Kernis, M. H., Greenier, K. D., Herlocker, C. E., Whisenhunt, C. R., & Abend, T.
A. (1997). Self-perceptions of reactions to doing well or poorly: The roles of
stability and level of self-esteem. Personality and Individual Differences, 22,
845-854.
Kernis, M. H., Lakey, C. E., & Heppner, W. L. (2008). Secure versus fragile high selfesteem as a predictor of verbal defensiveness: Converging findings across
three different markers. Journal of Personality, 76, 477-512.
Leary, M. R., Gallagher, B., Fors, E., Buttermore, N., Baldwin, E., Kennedy, K., &
Mills, A. (2003). The invalidity of disclaimers about the effects of social feedback on self-esteem. Personality and Social Psychology Bulletin, 29, 623-636.
Leary, M. R., & Downs, D. L. (1995). Interpersonal functions of the self-esteem motive: The self-esteem system as a sociometer. In M. Kernis (Ed.), Efficacy, agency, and self-esteem (pp. 123-144). New York: Plenum.
McClelland, G. H., & Judd, C. M. (1993). Statistical difficulties of detecting interaction and moderator effects. Psychological Bulletin, 114, 376-390.
Metalsky, G. I., Joiner, T. E., Hardin, T. S., & Abramson, L. Y. (1993). Depressive
reactions to failure in a naturalistic setting: A test of the hopelessness and selfesteem theories of depression. Journal of Abnormal Psychology, 102, 101-109.
Orth, U., Robins, R. W., & Meier, L. L. (2009). Disentangling the effects of low selfesteem and stressful events on depression: Findings from three longitudinal
studies. Journal of Personality and Social Psychology, 97, 307-321.
Paradise, A. W., & Kernis, M. H. (1999). Development of the contingent self-esteem scale.
Unpublished data, University of Georgia.
Park, L. E. (2010). Responses to self-threat: Linking self and relational constructs
with approach and avoidance motivation. Social and Personality Psychology
Compass, 4, 201-221.
Park, L. E., & Crocker, J. (2005). Interpersonal consequences of seeking self-esteem.
Personality and Social Psychology Bulletin, 31, 1587-1598.
Park, L. E., & Crocker, J. (2008). Contingencies of self-worth and responses to negative interpersonal feedback. Self and Identity, 7, 184-203.
Park, L. E., Crocker, J., & Kiefer, A. K. (2007). Contingencies of self-worth, academic failure, and goal pursuit. Personality and Social Psychology Bulletin, 33,
1503-1517.
Park, L. E., & Maner, J. K. (2009). Does self-threat promote social connection? The
role of self-esteem and contingencies of self-worth. Journal of Personality and
Social Psychology, 96, 203-217.
Pyszczynski, T., Greenberg, J., Solomon, S., Arndt, J., & Schimel, J. (2004). Why do
people need self-esteem? A theoretical and empirical review. Psychological
Bulletin, 130, 435-468.
1096ZEIGLER-HILL ET AL.
Roberts, J. E. (2006). Self-esteem from a clinical perspective. In M. H. Kernis (Ed.),
Self-esteem issues and answers: A sourcebook of current perspectives (pp. 298-305).
New York: Psychology Press.
Roberts, J. E., & Monroe, S. M. (1994). A multidimensional model of self-esteem in
depression. Clinical Psychology Review, 14, 161-181.
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton
University Press.
Spielberger, C. D. (1999). State-trait anger expression inventory manual. Odessa, FL:
Psychological Assessment Resources.
Vohs, K. D., & Heatherton, T. F. (2004). Ego threats elicit different social comparison
processes among high and low self-esteem people: Implications for interpersonal perceptions. Social Cognition, 22, 168-191.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief
measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54, 1063-1070.
Webster, G. D., Kirkpatrick, L. A., Nezlek, J. B., Smith, C. V., & Paddock, E. L. (2007).
Different slopes for different folks: Self-esteem instability and gender as moderators of the relationship between self-esteem and attitudinal aggression.
Self and Identity, 6, 74-94.
Wells, G. L., & Windschitl, P. D. (1999). Stimulus sampling and social psychological
experimentation. Personality and Social Psychology Bulletin, 25, 1115-1125.
Zeigler-Hill, V. (2011). The connections between self-esteem and psychopathology.
Journal of Contemporary Psychotherapy, 41, 157-164.
Zeigler-Hill, V., Clark, C. B., & Beckman, T. E. (in press). Fragile self-esteem and the
interpersonal circumplex: Are feelings of self-worth associated with interpersonal style? Self and Identity.